Supervised learning with a deep convolutional neural network is used to identify the QCD
equation of state (EoS) employed in relativistic hydrodynamic simulations of heavy-ion
collisions from the simulated final-state particle spectra ρ(pT,Φ). High-level correlations
of ρ(pT,Φ) learned by the neural network act as an effective "EoS-meter" in detecting
the nature of the QCD transition. The EoS-meter is model independent and insensitive to
other simulation inputs, especially the initial conditions. Thus it provides a powerful
direct-connection of heavy-ion collision observables with the bulk properties of QCD.
The quark-gluon plasma created in ultra-relativistic heavy-ion collisions behaves novelly
as a strongly coupled liquid, rather than a weakly coupled gas predicted theoretically.
This has been a key challenge for the theory community since the start of the RHIC
experiments in the year 2000, and long-range correlations are crucial missing ingredients
for a systematic approach based on the underlying theory of quantum chromo dynamics
(QCD). In this talk, I discuss a novel way to incorporate long-range correlations to thermal
QCD by systematically including the (chromo) magnetic scale g^2T — which is missing in
conventional approaches — in the setup. As a result, a novel massless mode is generated
by the magnetic scale which are long-range in nature, while the conventional massive
quasiparticle modes generated by the (chromo) electric scale gT are reproduced. The
residue of this massless mode is nonpositive at all temperatures, which consequently gives
rise to positivity violation in the quark spectral functions. This demonstrates profound
impacts of confinement effects on thermal quark collective excitations, which manifest
genuine long-range correlations in the system.
报告人简介：
1997.09-2003.07 中国科学技术大学 工程学学士
中国·浙江 湖州市二环东路759号（313000） 浙ICP备10025412号 
浙公网安备 33050202000195号 版权所有：党委宣传部
